A collaborative semantic framework based on activities for the development of applications in Smart Home living labs

Abstract

An essential element to improve the quality of life at home is to understand the behavior of its inhabitants and the activities they perform on a daily basis. The collection and analysis of this information in Smart Home environments poses several challenges, mainly data variability, and the management and synchronization of devices, which have limited the adoption of this kind of solutions. Researchers tend to work focused on providing solutions to specific issues and, in research environments that share infrastructure, such as living labs, it is common that previous developments of other teams are not reused. This process is suboptimal and does not allow understanding the Smart Home as a single space, but as a set of functional islands. To overcome these obstacles, a novel collaborative semantic framework is proposed for the extraction, characterization, and formalization of data to enable its reutilization and interpretation. The framework is based on the concept of activity, decomposed as a series of actions and events, and follows current trends in the domain of Activities of Daily Living and Smart Homes. A set of interoperability enablers based on JSON schemas and formalized descriptions of the domain have been proposed as the main facilitators of collaboration and the integration of elements coexisting in the Smart Home environment. The evaluation of the framework through the development of two research projects showed significant benefits, improving the overall quality of the solutions and enriching the extraction of insights through the correlated analysis of data.